Explainer: How 3D Gaussian Splatting Gives Robots Spatial Memory
TheTuringPost · x · 2026-10-22
3D Gaussian Splatting (3DGS) turns photos/videos captured by a robot into photorealistic 3D reconstructions viewable from new angles, giving robots a visual and spatial foundation for memory—letting them refer back to places and objects no longer in view.
How photos become a 3D scene:
- Camera pose estimation: match details across images to estimate camera positions and build a rough 3D point map;
- Gaussian initialization: points become tiny soft blobs, each with position, shape, color and opacity;
- Differentiable rendering optimization: rendered views are compared against photos; Gaussians shift, stretch and recolor until views match, with adaptive density control splitting or removing them to capture fine detail.
Once trained, the renderer enables real-time navigation across a wide variety of scenes.
More from Embodied
- LAGEN: one frame of latency can cut real-time agent performance by 95% — wzenus · 2026-10-22
- Oopsie Data: multi-lab effort to build first large-scale robot failure dataset — micoolcho · 2026-10-22
- Humanoid robot fights spark debate: entertainment is the only real use case today — KyleMorgenstein · 2026-10-22
- Unitree H2X humanoid robot sports dual battery slots in hands-on shots — micoolcho · 2026-10-22
- Unitree developer day draws 3,000 attendees as its robotics ecosystem takes shape — micoolcho · 2026-10-22
- Unitree's New H2-X Humanoid Pictured Next to G1, Height Gap Revealed — micoolcho · 2026-10-22